- Devel - libvirt List Archives

[libvirt] [PATCH] virsh: support up to 64 migration options for command
by Nikolay Shirokovskiy 14 Apr '16

14 Apr '16

Upcoming compression options for migration command patch series hits current limit of 32 possible options for a command. Lets take one step further and support 64 possible options. And all it takes is moving from 32 bit integers to 64 bit ones. The only less then trivial change i found is moving from 'ffs' to 'ffsl'. Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy(a)virtuozzo.com> --- tools/vsh.c | 29 ++++++++++++++--------------- 1 file changed, 14 insertions(+), 15 deletions(-) diff --git a/tools/vsh.c b/tools/vsh.c index 6bdc082..5659110 100644 --- a/tools/vsh.c +++ b/tools/vsh.c @@ -40,7 +40,6 @@ #include <limits.h> #include <sys/stat.h> #include <inttypes.h> -#include <strings.h> #include <signal.h> #if WITH_READLINE @@ -329,8 +328,8 @@ vshCmddefGetInfo(const vshCmdDef * cmd, const char *name) /* Validate that the options associated with cmd can be parsed. */ static int -vshCmddefOptParse(const vshCmdDef *cmd, uint32_t *opts_need_arg, - uint32_t *opts_required) +vshCmddefOptParse(const vshCmdDef *cmd, uint64_t *opts_need_arg, + uint64_t *opts_required) { size_t i; bool optional = false; @@ -344,7 +343,7 @@ vshCmddefOptParse(const vshCmdDef *cmd, uint32_t *opts_need_arg, for (i = 0; cmd->opts[i].name; i++) { const vshCmdOptDef *opt = &cmd->opts[i]; - if (i > 31) + if (i > 63) return -1; /* too many options */ if (opt->type == VSH_OT_BOOL) { optional = true; @@ -407,7 +406,7 @@ static vshCmdOptDef helpopt = { }; static const vshCmdOptDef * vshCmddefGetOption(vshControl *ctl, const vshCmdDef *cmd, const char *name, - uint32_t *opts_seen, int *opt_index, char **optstr) + uint64_t *opts_seen, int *opt_index, char **optstr) { size_t i; const vshCmdOptDef *ret = NULL; @@ -464,8 +463,8 @@ vshCmddefGetOption(vshControl *ctl, const vshCmdDef *cmd, const char *name, } static const vshCmdOptDef * -vshCmddefGetData(const vshCmdDef *cmd, uint32_t *opts_need_arg, - uint32_t *opts_seen) +vshCmddefGetData(const vshCmdDef *cmd, uint64_t *opts_need_arg, + uint64_t *opts_seen) { size_t i; const vshCmdOptDef *opt; @@ -474,7 +473,7 @@ vshCmddefGetData(const vshCmdDef *cmd, uint32_t *opts_need_arg, return NULL; /* Grab least-significant set bit */ - i = ffs(*opts_need_arg) - 1; + i = ffsl(*opts_need_arg) - 1; opt = &cmd->opts[i]; if (opt->type != VSH_OT_ARGV) *opts_need_arg &= ~(1 << i); @@ -486,8 +485,8 @@ vshCmddefGetData(const vshCmdDef *cmd, uint32_t *opts_need_arg, * Checks for required options */ static int -vshCommandCheckOpts(vshControl *ctl, const vshCmd *cmd, uint32_t opts_required, - uint32_t opts_seen) +vshCommandCheckOpts(vshControl *ctl, const vshCmd *cmd, uint64_t opts_required, + uint64_t opts_seen) { const vshCmdDef *def = cmd->def; size_t i; @@ -598,8 +597,8 @@ vshCmddefHelp(vshControl *ctl, const char *cmdname) const char *desc = vshCmddefGetInfo(def, "desc"); const char *help = _(vshCmddefGetInfo(def, "help")); char buf[256]; - uint32_t opts_need_arg; - uint32_t opts_required; + uint64_t opts_need_arg; + uint64_t opts_required; bool shortopt = false; /* true if 'arg' works instead of '--opt arg' */ if (vshCmddefOptParse(def, &opts_need_arg, &opts_required)) { @@ -1350,9 +1349,9 @@ vshCommandParse(vshControl *ctl, vshCommandParser *parser) const vshCmdDef *cmd = NULL; vshCommandToken tk; bool data_only = false; - uint32_t opts_need_arg = 0; - uint32_t opts_required = 0; - uint32_t opts_seen = 0; + uint64_t opts_need_arg = 0; + uint64_t opts_required = 0; + uint64_t opts_seen = 0; first = NULL; -- 1.8.3.1

2 1

[libvirt] [PATCH] virsh: use the same connection URI for reconnect
by Roman Bogorodskiy 14 Apr '16

14 Apr '16

When for some reason virsh looses connection and then tries to reconnection, it uses the default URI instead of the one that was used for the previous connection it got disconnected from. In order to make it reconnect using the same URI, copy URI of the current (disconnected) connection to vshControl 'connname' attribute. --- tools/virsh.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/tools/virsh.c b/tools/virsh.c index 5a61189..07097aa 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -359,6 +359,14 @@ virshConnectionHandler(vshControl *ctl) { virshControlPtr priv = ctl->privData; + if (disconnected && priv->conn) { + if (!ctl->connname) { + char *uri = virConnectGetURI(priv->conn); + ctl->connname = vshStrdup(ctl, uri); + VIR_FREE(uri); + } + } + if (!priv->conn || disconnected) virshReconnect(ctl); -- 2.7.4

2 3

[libvirt] [PATCH REBASE 0/4] vz: use disk bus and target name as disk id
by Nikolay Shirokovskiy 14 Apr '16

14 Apr '16

This series is a preparation to implement device update functionality. In case of disk updates disk bus and disk target name pair is considered disk id. We already have places in vz driver code where we use disk source as id. These places are either not critical (see boot patch) or correct (see detach patch) so we need not to fix them immediately to use more rubust id. But as this id will be introduced anyway for disk updates let's move to new id in existing code. This way we have single id for disks. Nikolay Shirokovskiy (4): vz: introduce vzsdk disk id function vz: fix detach disk to use new disk id vz: fix boot check to use new disk id vz: cleanup: remove trivial function src/vz/vz_sdk.c | 251 ++++++++++++++++++++++++++++---------------------------- 1 file changed, 126 insertions(+), 125 deletions(-) -- 1.8.3.1

1 4

[libvirt] [PATCH] storage: mpath: Don't error on target_type=NULL
by Cole Robinson 14 Apr '16

14 Apr '16

We use device-mapper to enumerate all dm devices, and filter out the list of multipath devices by checking the target_type string name. The code however cancels all scanning if we encounter target_type=NULL I don't know how to reproduce that situation, but a user was hitting it in their setup, and inspecting the lvm2/device-mapper code shows many places where !target_type is explicitly ignored and processing continues on to the next device. So I think we should do the same https://bugzilla.redhat.com/show_bug.cgi?id=1069317 --- src/storage/storage_backend_mpath.c | 7 +------ 1 file changed, 1 insertion(+), 6 deletions(-) diff --git a/src/storage/storage_backend_mpath.c b/src/storage/storage_backend_mpath.c index b5b4bb6..bf1b1ab 100644 --- a/src/storage/storage_backend_mpath.c +++ b/src/storage/storage_backend_mpath.c @@ -114,12 +114,7 @@ virStorageBackendIsMultipath(const char *dev_name) dm_get_next_target(dmt, next, &start, &length, &target_type, &params); - if (target_type == NULL) { - ret = -1; - goto out; - } - - if (STREQ(target_type, "multipath")) + if (STREQ_NULLABLE(target_type, "multipath")) ret = 1; out: -- 2.7.3

2 1

[libvirt] may be bug in my own commit?
by Vasiliy Tolstov 14 Apr '16

14 Apr '16

Hi. I'm try to backport some my changes to centos libvirt (1.2.17), and see that may be i miss IFA_ADDRESS when peerData is null, can somebody recheck me? https://libvirt.org/git/?p=libvirt.git;a=blobdiff;f=src/util/virnetdev.c;h=… - if (nla_put(nlmsg, IFA_ADDRESS, addrDataLen, addrData) < 0) - goto buffer_too_small; + if (peerData) { + if (nla_put(nlmsg, IFA_ADDRESS, addrDataLen, peerData) < 0) + goto buffer_too_small; + } This misses else i think.. -- Vasiliy Tolstov, e-mail: v.tolstov(a)selfip.ru

1 0

[libvirt] [PATCH] fix build by correcting functions order and src/Makefile.am
by Maxim Nestratov 14 Apr '16

14 Apr '16

commit 30c61901 added new functions to libvirt_private.syms not alpabetically sorted and erroneously added vz sources to STATEFUL_DRIVER_SOURCE_FILES, which triggered check-aclrules running while vz driver isn't ready for it yet. Pushing under build-breaker rule. --- src/Makefile.am | 1 - src/libvirt_private.syms | 12 ++++++------ 2 files changed, 6 insertions(+), 7 deletions(-) diff --git a/src/Makefile.am b/src/Makefile.am index 08ff301..f5e57c0 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -630,7 +630,6 @@ DRIVER_SOURCE_FILES = \ $(NULL) STATEFUL_DRIVER_SOURCE_FILES = \ - $(VZ_DRIVER_SOURCES) \ $(BHYVE_DRIVER_SOURCES) \ $(INTERFACE_DRIVER_SOURCES) \ $(LIBXL_DRIVER_SOURCES) \ diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index 1bcb3af..a9025f5 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -916,6 +916,11 @@ virCPUx86MakeData; # datatypes.h virConnectClass; +virConnectCloseCallbackDataCall; +virConnectCloseCallbackDataClass; +virConnectCloseCallbackDataGetCallback; +virConnectCloseCallbackDataRegister; +virConnectCloseCallbackDataUnregister; virDomainClass; virDomainSnapshotClass; virGetConnect; @@ -929,20 +934,15 @@ virGetSecret; virGetStoragePool; virGetStorageVol; virGetStream; -virConnectCloseCallbackDataGetCallback; -virNewConnectCloseCallbackData; -virConnectCloseCallbackDataUnregister; -virConnectCloseCallbackDataRegister; -virConnectCloseCallbackDataCall; virInterfaceClass; virNetworkClass; +virNewConnectCloseCallbackData; virNodeDeviceClass; virNWFilterClass; virSecretClass; virStoragePoolClass; virStorageVolClass; virStreamClass; -virConnectCloseCallbackDataClass; # fdstream.h -- 2.4.3

2 2

Re: [libvirt] [libvirt-users] Libvirtd running as root tries to access oneadmin (OpenNebula) NFS mount but throws: error: can’t canonicalize path
by TomK 14 Apr '16

14 Apr '16

Adding in libvir-list. Cheers, Tom K. ------------------------------------------------------------------------------------- Mobile: 416 618 8456 Home: 905 857 9652 Living on earth is expensive, but it includes a free trip around the sun. On 4/7/2016 7:32 PM, TomK wrote: > Hey All, > > I've an issue where libvirtd tries to access an NFS mount but errors > out with: can't canonicalize path '/var/lib/one//datastores/0 . The > unprevilidged user is able to read/write fine to the share. > root_squash is used and for security reasons no_root_squash cannot be > used. > > On the controller and node SELinux is disabled. > > [oneadmin@mdskvm-p01 ~]$ virsh -d 1 --connect qemu:///system create > /var/lib/one//datastores/0/38/deployment.0 > create: file(optdata): /var/lib/one//datastores/0/38/deployment.0 > error: Failed to create domain from > /var/lib/one//datastores/0/38/deployment.0 > error: can't canonicalize path '/var/lib/one//datastores/0/38/disk.1': > Permission denied > > I added some debug flags to get more info and added -x to the deploy > script. Closest I get to more details is this: > > 2016-04-06 04:15:35.945+0000: 14072: debug : > virStorageFileBackendFileInit:1441 : initializing FS storage file > 0x7f6aa4009000 (file:/var/lib/one//datastores/0/38/disk.1)[9869:9869] > 2016-04-06 04:15:35.954+0000: 14072: error : > virStorageFileBackendFileGetUniqueIdentifier:1523 : can't canonicalize > path '/var/lib/one//datastores/0/38/disk.1': > > https://www.redhat.com/archives/libvir-list/2014-May/msg00194.html > > Comment is: "The current implementation works for local > storage only and returns the canonical path of the volume." > > But it seems the logic is applied to NFS mounts. Perhaps it shouldn't > be? Anyway to get around this problem? This is CentOS 7 . > > Cheers, > Tom K. > ------------------------------------------------------------------------------------- > > Living on earth is expensive, but it includes a free trip around the sun. > > _______________________________________________ > libvirt-users mailing list > libvirt-users(a)redhat.com > https://www.redhat.com/mailman/listinfo/libvirt-users

3 18

[libvirt] [PATCH V5] libxl: support creating domain with VF assignment from a pool
by Chunyan Liu 14 Apr '16

14 Apr '16

Add codes to support creating domain with network defition of assigning SRIOV VF from a pool. Signed-off-by: Chunyan Liu <cyliu(a)suse.com> Signed-off-by: Jim Fehlig <jfehlig(a)suse.com> --- Rebase and send a new version. src/libxl/libxl_domain.c | 48 ++++++++++++++++++++++++++++++++++++++++++++++++ tests/Makefile.am | 3 +++ 2 files changed, 51 insertions(+) diff --git a/src/libxl/libxl_domain.c b/src/libxl/libxl_domain.c index 04962a0..bd5deaa 100644 --- a/src/libxl/libxl_domain.c +++ b/src/libxl/libxl_domain.c @@ -36,6 +36,7 @@ #include "virtime.h" #include "locking/domain_lock.h" #include "xen_common.h" +#include "network/bridge_driver.h" #define VIR_FROM_THIS VIR_FROM_LIBXL @@ -764,6 +765,10 @@ libxlDomainCleanup(libxlDriverPrivatePtr driver, if (net->ifname && STRPREFIX(net->ifname, LIBXL_GENERATED_PREFIX_XEN)) VIR_FREE(net->ifname); + + /* cleanup actual device */ + virDomainNetRemoveHostdev(vm->def, net); + networkReleaseActualDevice(vm->def, net); } } @@ -900,6 +905,46 @@ libxlDomainFreeMem(libxl_ctx *ctx, libxl_domain_config *d_config) return -1; } +static int +libxlNetworkPrepareDevices(virDomainDefPtr def) +{ + size_t i; + + for (i = 0; i < def->nnets; i++) { + virDomainNetDefPtr net = def->nets[i]; + int actualType; + + /* If appropriate, grab a physical device from the configured + * network's pool of devices, or resolve bridge device name + * to the one defined in the network definition. + */ + if (networkAllocateActualDevice(def, net) < 0) + return -1; + + actualType = virDomainNetGetActualType(net); + if (actualType == VIR_DOMAIN_NET_TYPE_HOSTDEV && + net->type == VIR_DOMAIN_NET_TYPE_NETWORK) { + /* Each type='hostdev' network device must also have a + * corresponding entry in the hostdevs array. For netdevs + * that are hardcoded as type='hostdev', this is already + * done by the parser, but for those allocated from a + * network / determined at runtime, we need to do it + * separately. + */ + virDomainHostdevDefPtr hostdev = virDomainNetGetActualHostdev(net); + virDomainHostdevSubsysPCIPtr pcisrc = &hostdev->source.subsys.u.pci; + + if (hostdev->mode == VIR_DOMAIN_HOSTDEV_MODE_SUBSYS && + hostdev->source.subsys.type == VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_PCI) + pcisrc->backend = VIR_DOMAIN_HOSTDEV_PCI_BACKEND_XEN; + + if (virDomainHostdevInsert(def, hostdev) < 0) + return -1; + } + } + return 0; +} + static void libxlConsoleCallback(libxl_ctx *ctx, libxl_event *ev, void *for_callback) { @@ -1050,6 +1095,9 @@ libxlDomainStart(libxlDriverPrivatePtr driver, virDomainObjPtr vm, goto cleanup; VIR_FREE(priv->lockState); + if (libxlNetworkPrepareDevices(vm->def) < 0) + goto cleanup_dom; + if (libxlBuildDomainConfig(driver->reservedGraphicsPorts, vm->def, cfg->ctx, &d_config) < 0) goto cleanup_dom; diff --git a/tests/Makefile.am b/tests/Makefile.am index b3f1144..db4f88b 100644 --- a/tests/Makefile.am +++ b/tests/Makefile.am @@ -534,6 +534,9 @@ endif ! WITH_XEN if WITH_LIBXL libxl_LDADDS = ../src/libvirt_driver_libxl_impl.la +if WITH_NETWORK +libxl_LDADDS += ../src/libvirt_driver_network_impl.la +endif WITH_NETWORK libxl_LDADDS += $(LDADDS) xlconfigtest_SOURCES = \ -- 1.8.5.6

2 1

[libvirt] [PATCH v4 0/6] vz: change vz driver to be stateful driver and other enhancements
by Maxim Nestratov 13 Apr '16

13 Apr '16

There is no benefit in providing two ways of connecting to vz driver: by connecting via daemon and directly from client. Both ways finally come to a host where vz daemon sits. Always connecting via daemon allows us to have a single list of domains and share it among all connections. Since v1: removed patch "z: remove close callback implementations" building fixed close callback functions are added to libvirt_private.syms reworked not to lose event subscribers when connections drop Since v2: removed "vz: change vzConnectIsAlive behavior" addressed mostly all comments on previous series changed "vz: build driver as module and don't register it on client's side" Since v3: removed privileged check in vzStateInitialize, we will not work in non- privileged mode anyway, while skipping initialization is not correct because cleanup function will be called anyway prlsdkNewDomainByHandle changed to check for domain presense in the list under lock Maxim Nestratov (6): vz: build driver as module and don't register it on client's side vz: introduce new vzDriver lockable structure and use it vz: fix possible vzDomainDefineXMLFlags and prlsdkNewDomainByHandle race vz: implement connectGetSysinfo hypervisor callback vz: remove vzDriverLock/Unlock function vz: minor cleanup daemon/Makefile.am | 4 + daemon/libvirtd.c | 9 ++ src/Makefile.am | 19 ++- src/libvirt.c | 7 - src/libvirt_private.syms | 7 + src/vz/vz_driver.c | 384 +++++++++++++++++++++++++++++------------------ src/vz/vz_sdk.c | 211 +++++++++++++------------- src/vz/vz_sdk.h | 30 ++-- src/vz/vz_utils.c | 27 ++-- src/vz/vz_utils.h | 34 +++-- 10 files changed, 431 insertions(+), 301 deletions(-) -- 2.4.3

3 13

[libvirt] [PATCH v2] vz: simplify getting strings from vz sdk
by Nikolay Shirokovskiy 13 Apr '16

13 Apr '16

changes from v1: ================ 1. move from macros to helper function as former is too heavy weigth for the task 2. squash fixed and variable length patches into one 3. rebase on master, so latest boot order patches can get benefits too Nikolay Shirokovskiy (1): vz: simplify getting strings from vzsdk src/vz/vz_sdk.c | 208 ++++++++++++++++++-------------------------------------- 1 file changed, 68 insertions(+), 140 deletions(-) -- 1.8.3.1

2 2